A few weeks ago, a research report by Oppenheimer started circulating in which the authors looked at the current state of the US longhaul transport market. In effect, they said that there is still a fiber glut and that there will continue to be a glut for another decade until there are only 3-4 players left in the longhaul market – with LVLT remaining the most active consolidator. I have been asked to comment on this report by a half dozen readers now, so here goes.
The whole premise here is that there is some point at which wholesale intercity transport will return to its bubble-era youth where fiber was perceived to be a scarce commodity you could turn into gold just by burying it. There isn’t, it won’t, and you gotta be kidding. In terms of raw bandwidth potential in the ground between major hubs, there will always and forever be a glut. Any interruption in that glut will be swiftly met with new fiber or new technology adding vast capacity and a new glut. Intercity fiber count remains so ubiquitous as to be irrelevant, as it still only takes one or two fiber pairs to field a top 10 IP transit backbone.
But actually, the concept of the fiber glut doesn’t matter any more in a practical sense. The industry has now learned how to make money off of fiber without expecting intercity wholesale longhaul to become a standalone healthy business model. It has been a very long time now since you could judge the health of the bandwidth industry by its least common demoninator. Raw capacity (and especially raw intercity fiber count) means nothing compared to the capabilities you enable, the services you offer, and the content you deliver. The fiber glut remains and may be eternal, but nobody cares anymore. In fact we’ve even seen new intercity fiber getting built on a few routes despite all that stuff in the ground already (Allied Fiber, Spread Networks).
Intercity fiber exists today as an enabler. Whether you lease waves, hold fiber IRUs, or built the conduit yourself, what it does for you is enable the rest of your product portfolio. For metro-focused infrastructure players, it connects up their markets and lets them serve regional customers all on-net. For enterprise-focused players, it helps maintain end-to-end control and keeps margins up. For cloud-focused players, it helps control latency and service quality.
Oppenheimer does note that longhaul services are increasingly bundled with metro, CDN, transit, colocation, managed services, etc and says that long-haul may cease to exist as a standalone business. But this train left the station years ago. Name one standalone longhaul network. Heck name one major longhaul network that gets more than half its revenues from intercity transport. Of the eight LD networks Oppenheimer discusses, precisely none of them is waiting around for the end of the ‘fiber’ glut.
There are a few points in the report that ought to be clarified a bit. Several times the authors say that Level 3 owns half of the fiber on the Qwest network (now CenturyLink’s), and once guesstimate it’s expense in the hundreds of millions of dollars per year in O&M. It is true that Level 3 acquired 24 of those fibers nationally each from Genuity and Global Crossing, plus others that WilTel and Broadwing had used to fill out their networks. However, I am pretty sure that they returned much of the Genuity fiber long ago, except those routes that they wanted to keep. I expect they will do the same with the Global Crossing footprint in reasonably short order. And it should also be noted that CenturyLink now owns several fibers on the original Level 3 network via the Savvis purchase, and pays Level 3 O&M for those.
There is also a bit of ambiguity about exactly where all this fiber glut resides, suggesting that the five or so largest players each hold a substantial hand. In fact, CenturyLink and Level 3 own nearly all of it now because they own the last two major national buildouts. The mostly older fiber builds of Sprint, Verizon, and AT&T simply do not match up to the newer national fiber footprints in terms of raw practically accessible capacity. There is a reason Verizon is pushing 100G so hard, and it’s not just because it simplifies their network. They are relatively fiber-poor in comparison from an intercity fiber point of view. Of the others cited, Cogent and AboveNet each have only a few strands while only XO is sitting on a relative bonanza of more than a dozen nationally. Is it still a glut when two parties own most of the theoretical capacity?
Not that it matters currently of course. As I was saying, what matters is what you do with the fiber you have, and on that front most of these longhaul networks is holding its own. There will certainly be consolidation from here of course, with the weakest hands being the targets in the short to medium term IMHO – currently XO and Sprint Wireline.
If you haven't already, please take our Reader Survey! Just 3 questions to help us better understand who is reading Telecom Ramblings so we can serve you better!Categories: Internet Backbones · Internet Traffic